A customer wants a scalable, repeatable way to roll out AI-assisted optimizations across many applications. What approach should the consultant recommend?
Select an answer to reveal the explanation.
Short Explanation
Scale doesn't change the discipline, it just means applying the same small-change, test-first pattern everywhere. You want each optimization reviewed and regression-tested before it merges, so the rollout stays repeatable across many apps.
Full Explanation
The correct approach is standardizing on applying optimizations as small, reviewable changes, each verified with regression tests before merging. A repeatable, scalable practice is the same discipline applied consistently: small changes, individually reviewed, backed by regression tests, so the process does not depend on any one application's size or complexity.
The approach of applying every suggested optimization from each analysis run in a single deployment per application is incorrect because bundling every suggestion into one deployment per application reintroduces the attribution problem at scale, across many applications at once.
The approach of skipping regression tests for optimization changes to keep the rollout timeline short is incorrect because dropping regression tests to save time removes the safeguard that makes the rollout safe to repeat.
The approach of deploying optimizations to production first and rolling back only if a customer reports an issue is incorrect because relying on customer reports to catch regressions means problems reach production before anyone notices, which is the opposite of a controlled rollout.
For the exam, the small-change-with-tests pattern is what scales; skipping it to move faster does not.